This program of research will continue its dual foci: 1) defining the phenotype of autism at multiple levels of analysis: behavior, neuropsychology, brain function, brain structure, and familial aspects; and 2) defining the continuities and discontinuities in the developmental course of autism using comparative longitudinal studies. In the first five years of funding, the investigators were able to eliminate or refine four of the competing neuropsychological explanations of autism: a basic sensory deficit, a praxis deficit, a cognitive intersubjectivity deficit, and an executive deficit. The investigators' past work has narrowed and deepened their focus to four key areas: 1) imitation, 2) core affective processes, 3) spatial working memory, and 4) relationships of brain structure and language functioning. In the proposed studies, they will examine these areas in autism again at multiple levels of analysis: behavior, neuropsychology, brain function and structure, and familial aspects. The investigators will conduct longitudinal studies comparing autism, fragile X syndrome, Down syndrome, other developmental delays, and typical development across ages spanning from early infancy (through the use of home videos) to middle childhood, and will also include familial data from parents. Research methods including behavioral and neuropsychological measures, brain imaging techniques including magnetoencephalography and magnetic resonance imaging, psychophysiological paradigms involving eye movements, electrodermal responses, electromyography, and comparison of data from children and their parents. The main aims of this program are: 1) to refine and delineate autism-specific impairments, spared abilities, and familial aspects of imitation, emotional processes, spatial working memory, and a relationship between language and brain structure; 2) to examine whether skills in these areas are independent or whether relationships among them exist concurrently and longitudinally; 3) to determine whether the patterns of deficits are more suggestive of homogeneity or heterogeneity (suggesting potential subtypes); and 4) to participate in cross-network projects that examine the biological, neuropsychological, and behavioral aspects of autism.